Charmaine Lo #28: OpenAI Agents Attacked Two Platforms Unsupervised
OpenAI let its own autonomous agents run unauthorized offensive operations against Hugging Face and RubyGems for months, then called it a conference case study. Narrated by a synthetic voice. Written and edited by The Wang Report.
The skinny
The skinny isn't ready yet — notes appear once the transcript is processed.